Résumé
Optimal transport (OT) theory focuses, among all maps T : ℝd → ℝd that can morph a probability measure onto another, on those that are the “thriftiest”, i.e. such that the averaged cost c(x, T(x)) between x and its image T(x) be as small as possible. Many computational approaches have been proposed to estimate such Monge maps when c is the ℓ22 distance, e.g., using entropic maps (Pooladian and Niles-Weed, 2021), or neural networks (Makkuva et al., 2020; Korotin et al., 2020). We propose a new model for transport maps, built on a family of translation invariant costs c(x, y):= h(x - y), where h := 1/2 ∥ · ∥2/2 + τ and τ is a regularizer. We propose a generalization of the entropic map suitable for h, and highlight a surprising link tying it with the Bregman centroids of the divergence Dh generated by h, and the proximal operator of τ. We show that choosing a sparsity-inducing norm for τ results in maps that apply Occam's razor to transport, in the sense that the displacement vectors ∆(x):= T(x) - x they induce are sparse, with a sparsity pattern that varies depending on x. We showcase the ability of our method to estimate meaningful OT maps for high-dimensional single-cell transcription data, in the 34000-d space of gene counts for cells, without using dimensionality reduction, thus retaining the ability to interpret all displacements at the gene level.
| langue originale | Anglais |
|---|---|
| Pages (de - à) | 6671-6682 |
| Nombre de pages | 12 |
| journal | Proceedings of Machine Learning Research |
| Volume | 202 |
| état | Publié - 1 janv. 2023 |
| Modification externe | Oui |
| Evénement | 40th International Conference on Machine Learning, ICML 2023 - Honolulu, États-Unis Durée: 23 juil. 2023 → 29 juil. 2023 |
Empreinte digitale
Examiner les sujets de recherche de « Monge, Bregman and Occam: Interpretable Optimal Transport in High-Dimensions with Feature-Sparse Maps ». Ensemble, ils forment une empreinte digitale unique.Contient cette citation
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver